# maal0lan/cyberbully architecture

> Architecture diagram of the GitHub repository maal0lan/cyberbully, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-02). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/maal0lan/cyberbully
- Repository: https://github.com/maal0lan/cyberbully

## Overview

The repository combines cyberbullying dataset creation and model development with a single-text detection workflow. The documented user path is text input → CLI prediction → label, probability, and threshold; the training script prepares data, trains, evaluates, and exports a model. Dataset-generation pipelines are represented as separate product capabilities. Some training and CLI internals were not excerpted, so those links rely on README statements rather than claims about unshown implementation.

Main entry points: [`dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/explicit%20generator/generate_pipeline.py) (Explicit generation), [`dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py`](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py) (Clean generation), [`predict_cyberbullying_cli.py`](https://github.com/maal0lan/cyberbully/blob/main/predict_cyberbullying_cli.py) (Prediction CLI), [`cyberbully_output/best_model.keras`](https://github.com/maal0lan/cyberbully/blob/main/cyberbully_output/best_model.keras) (Model checkpoint)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_generation["Dataset generation"]
  node_explicitgen["Explicit generation"]
  node_explicittemplates["Explicit prompts<br/>[templates.py]"]
  node_clean_gen["Clean generation<br/>[pipeline_clean.py]"]
  node_clean_templates["Clean prompts<br/>[templates_clean.py]"]
  node_sentence_maker["Sentence generation<br/>[sentence_maker.py]"]
  node_language_names["Language file renaming<br/>[name_changer.py]"]
end

subgraph group_curation["Dataset curation"]
  node_augment_explicit["Explicit augmentation"]
  node_augment_clean["Clean augmentation"]
  node_merge["Dataset merger<br/>[merge_datasets.py]"]
  node_merged_data[("Merged dataset")]
  node_analytics["Dataset analytics"]
end

subgraph group_model["Model workflow"]
  node_trainer["Training pipeline"]
  node_prepared["Prepared splits"]
  node_checkpoint["Model checkpoint<br/>[best_model.keras]"]
end

subgraph group_serving["Prediction and evaluation"]
  node_cli["Prediction CLI"]
  node_prediction["Label and score"]
  node_evaluation["Model evaluation<br/>[evaluate_models.py]"]
  node_evalresults["Evaluation results"]
end

node_user(("Text submitter"))
node_ollama["Ollama model service"]

node_user -->|"submits text"| node_cli
node_cli -->|"loads model"| node_checkpoint
node_cli -->|"prints prediction"| node_prediction
node_trainer -->|"creates splits"| node_prepared
node_trainer -->|"exports model"| node_checkpoint
node_evaluation -->|"evaluates model"| node_checkpoint
node_evaluation -->|"writes metrics"| node_evalresults
node_explicitgen -->|"uses prompts"| node_explicittemplates
node_explicitgen -->|"requests generation"| node_ollama
node_explicitgen -->|"requests judging"| node_ollama
node_explicitgen -->|"filters agreements"| node_explicitgen
node_clean_gen -->|"uses prompts"| node_clean_templates
node_clean_gen -->|"requests generation"| node_ollama
node_clean_gen -->|"requests judging"| node_ollama
node_augment_explicit -->|"mutates source words"| node_augment_explicit
node_augment_clean -->|"mutates source words"| node_augment_clean
node_merge -->|"writes merged data"| node_merged_data
node_analytics -.->|"analyzes dataset"| node_merged_data

click node_cli "https://github.com/maal0lan/cyberbully/blob/main/predict_cyberbullying_cli.py"
click node_trainer "https://github.com/maal0lan/cyberbully/blob/main/cyberbully_final.py"
click node_prepared "https://github.com/maal0lan/cyberbully/blob/main/cyberbully_final.py"
click node_checkpoint "https://github.com/maal0lan/cyberbully/blob/main/cyberbully_output/best_model.keras"
click node_evaluation "https://github.com/maal0lan/cyberbully/blob/main/eval_results/evaluate_models.py"
click node_evalresults "https://github.com/maal0lan/cyberbully/tree/main/eval_results"
click node_explicitgen "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/explicit%20generator/generate_pipeline.py"
click node_explicittemplates "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/explicit%20generator/templates.py"
click node_clean_gen "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py"
click node_clean_templates "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/non-explicit-generator/templates_clean.py"
click node_augment_explicit "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/explicit_dataset_generation/exp_augment_pipeline.py"
click node_augment_clean "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/non-explicit_dataset_generation/non-exp_augment_pipeline.py"
click node_merge "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/merge_datasets.py"
click node_merged_data "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/dataset/generated_dataset/cyberbullying_dataset.csv"
click node_analytics "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/dataset/analytics/dataset_analytics.py"
click node_sentence_maker "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/data_words/sentence_maker.py"
click node_language_names "https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/data_words/name_changer.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_explicitgen,node_explicittemplates,node_clean_gen,node_clean_templates,node_sentence_maker,node_language_names toneBlue
class node_augment_explicit,node_augment_clean,node_merge,node_merged_data,node_analytics toneAmber
class node_trainer,node_prepared,node_checkpoint toneMint
class node_cli,node_prediction,node_evaluation,node_evalresults toneRose
class node_user,node_ollama toneIndigo
```

## Components

### Dataset generation

- [Explicit generation](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/explicit%20generator/generate_pipeline.py): `dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`
- [Explicit prompts](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/explicit%20generator/templates.py): `dataset_generation/helper_files/gen/explicit generator/templates.py`
- [Clean generation](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py): `dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py`
- [Clean prompts](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/gen/non-explicit-generator/templates_clean.py): `dataset_generation/helper_files/gen/non-explicit-generator/templates_clean.py`
- [Sentence generation](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/data_words/sentence_maker.py): `dataset_generation/data_words/sentence_maker.py`
- [Language file renaming](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/data_words/name_changer.py): `dataset_generation/data_words/name_changer.py`

### Dataset curation

- [Explicit augmentation](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/explicit_dataset_generation/exp_augment_pipeline.py): `dataset_generation/helper_files/expander/explicit_dataset_generation/exp_augment_pipeline.py`
- [Clean augmentation](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/non-explicit_dataset_generation/non-exp_augment_pipeline.py): `dataset_generation/helper_files/expander/non-explicit_dataset_generation/non-exp_augment_pipeline.py`
- [Dataset merger](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/expander/merge_datasets.py): `dataset_generation/helper_files/expander/merge_datasets.py`
- [Merged dataset](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/dataset/generated_dataset/cyberbullying_dataset.csv): `dataset_generation/helper_files/dataset/generated_dataset/cyberbullying_dataset.csv`
- [Dataset analytics](https://github.com/maal0lan/cyberbully/blob/main/dataset_generation/helper_files/dataset/analytics/dataset_analytics.py): `dataset_generation/helper_files/dataset/analytics/dataset_analytics.py`

### Model workflow

- [Training pipeline](https://github.com/maal0lan/cyberbully/blob/main/cyberbully_final.py): `cyberbully_final.py`
- [Prepared splits](https://github.com/maal0lan/cyberbully/blob/main/cyberbully_final.py): `cyberbully_final.py`
- [Model checkpoint](https://github.com/maal0lan/cyberbully/blob/main/cyberbully_output/best_model.keras): `cyberbully_output/best_model.keras`

### Prediction and evaluation

- [Prediction CLI](https://github.com/maal0lan/cyberbully/blob/main/predict_cyberbullying_cli.py): `predict_cyberbullying_cli.py`
- **Label and score**
- [Model evaluation](https://github.com/maal0lan/cyberbully/blob/main/eval_results/evaluate_models.py): `eval_results/evaluate_models.py`
- [Evaluation results](https://github.com/maal0lan/cyberbully/tree/main/eval_results): `eval_results`

### Other components

- **Text submitter**
- **Ollama model service**

## Connections

- Text submitter → Prediction CLI: submits text (evidence: `README.md`)
- Prediction CLI → Model checkpoint: loads model (evidence: `README.md`)
- Prediction CLI → Label and score: prints prediction (evidence: `README.md`)
- Training pipeline → Prepared splits: creates splits (evidence: `cyberbully_final.py`)
- Training pipeline → Model checkpoint: exports model (evidence: `README.md`)
- Model evaluation → Model checkpoint: evaluates model (evidence: `README.md`)
- Model evaluation → Evaluation results: writes metrics (evidence: `eval_results/evaluate_models.py`)
- Explicit generation → Explicit prompts: uses prompts (evidence: `dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`)
- Explicit generation → Ollama model service: requests generation (evidence: `dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`)
- Explicit generation → Ollama model service: requests judging (evidence: `dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`)
- Explicit generation → Explicit generation: filters agreements (evidence: `dataset_generation/helper_files/gen/explicit generator/generate_pipeline.py`)
- Clean generation → Clean prompts: uses prompts (evidence: `dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py`)
- Clean generation → Ollama model service: requests generation (evidence: `dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py`)
- Clean generation → Ollama model service: requests judging (evidence: `dataset_generation/helper_files/gen/non-explicit-generator/pipeline_clean.py`)
- Explicit augmentation → Explicit augmentation: mutates source words (evidence: `dataset_generation/helper_files/expander/explicit_dataset_generation/exp_augment_pipeline.py`)
- Clean augmentation → Clean augmentation: mutates source words (evidence: `dataset_generation/helper_files/expander/non-explicit_dataset_generation/non-exp_augment_pipeline.py`)
- Dataset merger → Merged dataset: writes merged data (evidence: `dataset_generation/helper_files/expander/merge_datasets.py`)
- Dataset analytics → Merged dataset: analyzes dataset
